The ATP-binding cassette (ABC) transporters such as P-glycoprotein (Pgp), the multidrug resistance-associated protein (MRP1), and the mitoxantrone-resistance protein (MXR also known as breast cancer resistance protein, BCRP, ABCP or ABCG2), which function as ATP-dependent efflux pumps, play an important role in the development of multidrug resistance in most cancers. There are 48 known ABC transport proteins in the human and at least 10 of these transporters are involved in the movement of a variety of amphipathic agents including anticancer agents, nucleotide analogs and cyclic nucleotides. Thus, some of these transporters also may contribute to the development of multidrug resistance in malignant cells. Our studies are directed toward understanding the mechanism of action of the multidrug resistance-linked ABC transporters. By using a baculovirus-insect cell expression system, a large amount (~5-10 mg) of biologically active Pgp has been prepared for biophysical and structural studies, as further understanding of the mechanism of these transporters would be accelerated by resolution of the structure of Pgp. In last couple of years we have directed our efforts towards understanding the catalytic cycle of ATP hydrolysis by Pgp, identification of rate-limiting step(s) and modulation of the ATPase activity by substrates and modulators. Similar studies have been initiated with MRP1, MRP4 and MRP8 to gain insight into the role of the two ATP sites in ATP hydrolysis by these transporters. Such studies will provide an insight into the role of these transporters in the development of multidrug resistance in cancers and aid in the development of new therapeutic strategies. 1. Elucidation of the catalytic cycle of ATP hydrolysis by Pgp: Recently, we have provided evidence for two distinct roles for ATP hydrolysis in a single turnover of Pgp, one in the transport of drug and the other in effecting conformational changes to reset the transporter for the next catalytic cycle. Our goal is to understand the role of two ATP sites of Pgp in ATP hydrolysis. Our recent work with mutations in the conserved Glu residues (E556 and E1201) in the Walker B domain of either the N-or C-terminal ATP site suggest that these residues are not required for the cleavage of the bond between b-P and g-P of ATP but are essential for the completion of catalytic cycle. We have postulated that these Glu residues are involved in transmission of signal from the substrate binding sites via the D-loop and signature region to the P-loop. We are testing this hypothesis by substituting the conserved residues in the D-loop, H-loop and the signature region with either Gln or Ala residues. In addition, we also find that replacement of a conserved Tyr residue (25 aa upstream of the Walker A domain) with Ala at position 401 (N-ATP site) or 1044 (C-ATP site) results in significant reduction in ATP binding. Y401A, Y1044A, and the Y401A/Y1044 double mutant could neither bind and trap nucleotide nor carry out steady-state ATP hydrolysis. These data suggest that although both Y401 and Y1044 are essential for ATP-binding and hydrolysis, the Tyr residue at this position in either ATP site can be replaced with another aromatic residue such as Phe or Trp without affecting function. However, the replacement of Tyr in both ATP sites simultaneously with Trp, Cys or Ala is not tolerated. Further studies of these mutants will help us understand the structural flexibility of Y401 and Y1044 residues and their role in the catalytic cycle of ATP hydrolysis by Pgp. 2. Characterization of substrate interaction sites and identification of regions and residues involved in interaction of substrates with Pgp: Our previous studies demonstrated the presence of at least two non-identical substrate interaction sites on Pgp. We have continued our efforts to characterize the interactions of substrates and inhibitors to elucidate the biochemical basis for the broad substrate specificity of Pgp. We have screened a large variety of derivatives of stipiamide, a synthetic polyene antibiotic that reverses Pgp-mediated drug resistance, to gain insight into the substrate interaction sites. We have also studied the mechanism by which disulfiram, a drug used to treat alcoholism, modulates Pgp activity. From a clinical perspective it is interesting that disulfiram also modulates activities of MRP1 and MRP4. We are focusing on three aspects: biochemical basis of the action of modulators; identification of regions and residues interacting with substrates/modulators; and biochemical characterization of substrate/modulators sites on yeast ABC transporters, Pdr5p and Cdr1p, which are functionally similar to mammalian Pgp. These data suggest the existence of at least three substrate-binding sites on Pdr5p that differ in behavior from those of the mammalian Pgp and some substrates appear to interact at more than one site. In addition, the observation that Pdr5p can confer resistance to tetraalkyltins suggests that one of the sites might use only hydrophobic interactions to bind substrates. We found that Cdr1p in purified plasma membrane fraction can be labeled with photoaffinity analogs of mammalian Pgp substrates, IAAP and azidopine. The labeling of Cdr1p with IAAP is inhibited by Nystatin in a dose-dependent manner, but this labeling is not affected by fluconazole, miconazole and cycloheximide. On the other hand, the azidopine labeling is inhibited by miconazole but not by Nystatin or cycloheximide. These findings suggest that IAAP and azidopine may be interacting at separate sites on Cdr1p. 3. Characterization of human MRP1 (ABCC1) and MRP8 (ABCC11): Multidrug resistance protein (MRP1), encoded by the gene MRP1 in humans, confers resistance to multiple natural product drugs by reducing the cellular drug concentration. We have optimized conditions for expression of MRP1 protein in recombinant vaccinia virus infected-transfected HeLa cells. We have also developed MRP1 transfected HEK293 and HeLa cells. The preliminary characterization of the catalytic cycle of ATP hydrolysis by MRP1 indicates that some of the features appear to be similar to those observed with Pgp. We are comparing the catalytic cycle of ATP hydrolysis by MRP1 with that of Pgp to understand the mechanistic differences between these two transporters. Recently, two new members of the MRP subfamily have been identified (MRP8 and MRP9). We have expressed human MRP8 in baculovirus-infected High Five insect cells. The initial results indicate that MRP8, similar to Pgp and MRP1, exhibits substrate-stimulated ATPase activity. We plan to characterize the MRP8 transporter using stably transfected cell lines and transient Vaccinia and baculovirus-based expression systems. Such studies will help us to understand the function of these MRP subfamily members and their role in the development of multidrug resistance in cancer cells. 4. Resolution of two- and three-dimensional structure of human Pgp: The high-resolution structure of Pgp at various stages during the catalytic cycle will be essential to understand the transport mechanism. This is one of our major interests and we have invested considerable effort in the past to develop methods for obtaining pure and active Pgp in large amounts. We have developed methods for large-scale purification of human Pgp expressed in baculovirus-MDR1 infected insect cells by metal affinity (Talon) and anion exchange (DE52) chromatography. Recently, by making improvements in our methods, we have obtained pure Pgp at ~6 to 7 mg/ml concentration. Most importantly, even at such a high concentration, Pgp in detergent solution is mainly present as a monomer.